Page History

Current Sprint - Sprint 21 ( January 07, 2016 – January 20, 2016 )

LexEVS 6.4 S13-500 Agile Development - Sprint StatusStatus#SprintStatus-Sprint21

Issue Prioritization

Recent LexEVS Related Bugs and Features (within last week)

Recent CTS2 Service Related Issues (within last week)

Issue Grooming

Prioritize JIRA issues (fix version 6.4)

Touching This discussion is to touch base with three different groups to bring everyone up to speed on the latest developments: Northrup Grummond, Mayo, and Gilberto.

Jason - Now has a "customer managed" VM set up.  Still need to move the existing triple store to this VM.

Jason's team also looked at NCBO OWL converter.  They had an issue with NCI OWL being converted (RRF to turtle).

Scott asked what is the use case for the triple store? They NCI suggested the following:

• To expose NCI thesaurus
• Expose a SPARQL endpoint
• Anything that will help the browser to display relationships. 

Another use case Next, would be for a triple store over the Metathesaurus.

Gilberto suggested two use cases:

• In general, a SPARQL endpoint could be used for better semantic web integration.  To facilitate queries across systems (federated queries) for the NCI thesaurus.
• Specific for LexEVS  LexEVS,  it should either support a general endpoint or integration integrate the functionality within our codebase.

Scott suggested that if a SPARQL endpoint is behind LexEVS, our hierarchy API (isA relationships) would call the endpoint instead of the underlying DB now.

• Larry mentioned that the endpoint may be used for other relationships than just the hierarchy API mentioned above.

Graph DB vs. Triple store - there hasn't been a decision to go with one or the other.  If a graph DB works better for LexEVs, than this would be an option.

Gilberto mentioned that the paper review is complete.  They will be getting demos from 4 vendors:

• Alegro Graph
• Stardog
• Virtuoso
  
  

CTRP Planning

• Update from Larry

Value Set Resolution

Tree Extension Issue

Scott reviewed this issue and this will need further investigation.  Depending on how critical this is, it may need to be looked at during the next Sprint.

Jason mentioned this is critical for the next release of the browser - release 2.8. 

The scenario is when you pull up a tree, and click on the "..." it freezes the whole tree.

Jason said they are planning to deploy in April.  We will need to prioritize this issue to be completed before then.  Larry suggested that we work on this right after the lucene issues are fixed.

Search Extension Testing

The search extension API is now complete.  We would like to understand the use cases from the NCI team. 

We will need to test 6.3 and 6.4 in a test harness to determine any performance issues.

In 6.3, search (query across a large number of systems) could take a long time (timeout after 5 minutes).